1. Field
The present disclosure relates generally to a data processing system and more particularly to data management. Still more particularly, the present disclosure relates to redundancy management in data processing systems.
2. Background
High-availability computing cluster solutions have historically focused on failure recovery, where upon failure of the primary service provider computer program, some time has to be spent transitioning to the backup service provider computer program. During this transition time, the critical service can be unavailable to requesting computer programs. Further, distribution of running instances of computer programs across a computing cluster should take into account the redundant characteristics and limitations of the cluster hardware so that the failure of any single hardware component does not render the service provided by the redundant computer programs unavailable. This distribution of instances should take into account the operable state of the cluster, that is, in the face of component failures the distribution of instances should use the remaining hosts respecting the residual redundancy.
Therefore, it would be advantageous to have a method and apparatus that addresses one or more of the issues discussed above.